6.4 Keynote - Horsley
-
Upload
krishna-kumar -
Category
Documents
-
view
224 -
download
0
Transcript of 6.4 Keynote - Horsley
-
8/6/2019 6.4 Keynote - Horsley
1/17
1
Improved Inspection and Assessment ofPipeline Welds
Advanced Welding and Joining Technical Workshop
David Horsley
Boulder COJanuary 25, 2006
2
2
Outline
Background Information
Problem Statement
Design Requirements
Review of welding, inspection and assessment technologies
Current
New
Future
Introduction to gaps and challenges
Weld Design
Inspection
Assessment
-
8/6/2019 6.4 Keynote - Horsley
2/17
-
8/6/2019 6.4 Keynote - Horsley
3/17
5
5
Current Technologies Widespread Implementation
WorkmanshipStandards
ECA Standards forStress Based Design
Radiography
Manual UTAutomated UT
Mag ParticleDye Penetrant
SMAW
FCAWMechanized GMAW
AssessmentInspectionWelding
These technologies are used worldwide.
6
6
New Technologies Limited / Recent Implementation
ECA Methods (oftencompany specific) for Severe
Loading
Strain
CyclicECA Standards for SevereLoading
DNV
(CSA Z662-07)
Automated UT usingPhased Arrays
Improved AUT Interface
Phased Arrays forsleeves and branch
connections
Multi-wire MechanizedGMAW
TandemDual Tandem
Advanced Welding SystemsPosition-basedParameters
Digital QA/QC
Communications
AssessmentInspectionWelding
Welding:
Advanced systems give much greater control of welding parameters, and provide
more information than was previously available.
Inspection:
Phased arrays offer the potential to account for variability not previously tolerated.
Modern software provides easy to visualize and interpret information to the
operator.
Assessment:
Current codes are not addressing industry needs (eg. API 1104 Appendix A).
Companies will often implement their own assessment methodologies to address
known deficiencies, particularly with respect to more severe conditions such asstrain based loads
-
8/6/2019 6.4 Keynote - Horsley
4/17
7
7
Future Technologies
Improved ECA Predictions
Reliability Based Designand Assessment
Enhanced Phased Array UT
Improved Visualization andInterpretation Interface
Intelligent Systems
Adaptive/IntelligentGMAW Welding Systems
LAZER Assisted GMAWStand Alone LAZER
Friction Stir Welding
AssessmentInspectionWelding
The previous presentation by Nate Ames showed several technologies that will
change the way that mechanized GMAW welding systems are/or will be controlled:
-Real-time adaptive control-Single sensor differential thermal anhalysis
-Audible noise
Phased array systems are at infancy stage of development. Technologies borrowed
from medical and geophysics/seismic industries will be adapted to pipeline weld
inspection. They have the potential to account for, and adapt to, many of the
problems we see in current systems (e.g. temperature, acoustic velocities,
misalignment, etc.) and give much better POD, resolution and accuracy, and do
much of the interpretation currently left to the UT operator.
As our modeling capabilities advance, and we extend our database of experimental
tests, we will have much deeper understanding of the influences of the various
parameters on the overall integrity of the welds. By developing tools specifically
tailored to reliability based methods will be key inputs to the overall RBDA
methodology. RBDA will provide a means to appropriately design pipelines to
ensure a minimum level of risk is achieved.
8
8
Design Considerations
Weld:
Pipe Properties
Process, Consumable, Procedure+
Weld / HAZ Properties
Inspection:
Probability of Detection
Accuracy and Precision
Assessment:
Variability of Loads and Properties
Accuracy of Prediction
Safety Factor
WeldDesign
Requirements
Flaws
AssessmentInspection
-
8/6/2019 6.4 Keynote - Horsley
5/17
9
9
Weld Flaw Considerations
Fitness for service depends on:
Weld/HAZ Properties
Pipe Properties
Loads
Flaws
All are Variable:
Weld/HAZ Strength and Toughness Procedure, fit-up,oclock, preheat & inter-pass temp., weld/weld
Pipe Strength and Toughness Heat to heat, oclock, along
length, pipe/pipe
Loads Installation, service conditions, oclock, location on
ROW
Flaws Length, height, depth, oclock
10
10
Procedure Variation
Single Torch Narrow Offset15.3 mm W.T.
Single Torch14.3 mm W.T.
Single Torch Tandem13.4 mm W.T.
-
8/6/2019 6.4 Keynote - Horsley
6/17
11
11
0
200
400
600
800
1000
1200
0 5 10 15 20 25 30
Strain (%)
Stress(MPa)
Westpath
Narrow OffsetRound 4 mm diameter
Strip 4 x 9 mm
Round 5 mm diameter
Strip 5 x 9 mm
Charpy Impact Energy
@ -10C = 132 J
@ -45C = 105 J
Charpy Impact Energy
@ -10C = 225 J
@ -45C = 204 J
Single Torch
Procedure Variation
This slide shows that very different properties can be achieved with only slight
changes to the width of bevel.
Slight differences are seen due to round as compared to rectangular specimen types.
12
12
Stress-Strain Curves: Round vs Strip
Single Torch Tandem X100 Weld
0
200
400
600
800
1000
1200
0 5 10 15 20 25 30
Strain (%)
Stress(MPa)
Round -3.75 mm
Strip -8 x 4 mm
Strip -6.3 x 4.6 mm
Strip -5.4 x 2.1 mm
This slide shows several test results from the same weld, but using different
specimen types.
-
8/6/2019 6.4 Keynote - Horsley
7/17
13
13
Stress-Strain Curves-Split Strip
Single Torch Tandem X100 Weld
0
100
200
300
400
500
600
700
800
900
1000
0 5 10 15 20 25
Strain (%)
Stress(MPa)
OD- Split Strip 4 x 4 mm MT- Split Strip 4 x 4 mm
M
T
T- Split Strip towards OD
M- Split Strip from mid-thickness
This slide shows differences in properties at different locations in the through-wall
direction.
The important message from these curves is that welding parameters, sample
location, and specimen type all have effects on the resulting stress strain curve.
Because material properties are a key input into flaw assessment, it is extremely
important to understand the effect of so-called essential variables.
14
14
Weld Challenges and Gaps
Challenges
What are the weld and HAZ properties?
How to measure properties?
How do these properties vary?
What causes the variation and how can it be controlled?
GAPS
Need standard procedures for measuring properties Need guidance on quantifying variation for a given
procedure
Need to understand essential variables with respect to
the level of assessment
-
8/6/2019 6.4 Keynote - Horsley
8/17
15
15
Inspection
Probability of Detection
Accuracy and Precision
Appropriate to Design Requirements
16
16
Automated Ultrasonic Inspection
In the context of Flaw Assessment - Inspection
System Must:
Detect flaws
Determine height
Determine length
Determine position in WT
-
8/6/2019 6.4 Keynote - Horsley
9/17
17
17
Ultrasonic Testing
Detection depends on reflection of sound waves
Typically use probes that focus sound on target area
Degree of focus depends on design of lens
Strength of reflected wave depends on many factors
Size of reflector (flaw)
Shape
Orientation
Position
System must accommodate:
Wide range of temperatures
Variable acoustic properties
18
18
Automated Ultrasonic Inspection
Two philosophies for sizing in common use:
1. Full Zone Height Assumption
2. Amplitude Based
Full Zone Height Assumption:
The weld bevel is divided into numerous zones, and each zone is interrogated by adedicated ultrasonic probe. If a signal stronger than the calibrated threshold is
detected on that probe, the flaw is assumed to be the full height of that particular
zone. Adjacent channels are examined to determine if the flaw is located across two
or more zones. And again, if a signal is present the flaw is assumed to lie within
that entire zone. No attempt is made to size the flaw on the basis of signal strength.
Amplitude based sizing:
The weld bevel is divided into zones (but typically fewer than described above).
Signals above threshold strength are interpreted, and an estimate of flaw height is
made on the basis of signal strength.
-
8/6/2019 6.4 Keynote - Horsley
10/17
19
19
Full Zone Height Assumption
45
5
37.5
ROOT
LCP
FILL1
HP1
HP2
FILL2
UpStream
DownStream
45
5
37.5
ROOT
LCP
FILL1
HP1
HP2
FILL2
UpStream
DownStream
-6.0
-4.0
-2.0
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0
-6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58
-6.0
-4.0
-2.0
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0
-6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58
Root H
P1
HP2
LCP
Cap
F2 F1
Weld bevel is divided into many zones
One UT channel assigned to each zone
Flaw is assumed to be the full height of
the zone if signal is above threshold
Advantage:
Easy interpretation
Disadvantage
Numerous probes required
20
20
Amplitude Based Sizing
Weld bevel is divided into zones
One UT channel assigned to each zone
Signal is manually interpreted to estimate flaw height
based on strength of signal
Advantage:
Fewer probes required
Disadvantage:
Operator dependent
Size uncertainty
Requires time for interpretation
-
8/6/2019 6.4 Keynote - Horsley
11/17
21
21
Planar Reflector Aimed back at Transducer
Reflected Beam
Flaw
TransmittedFocused
Beam
Transducer
For Illustration Only
This cartoon is for illustration only. It illustrates a flaw located at, and slightly
smaller than, the target area. The flaw is a good reflector, and therefore a good
strong signal is received back at the transducer.
22
22
Convex Reflector
TransmittedFocused Beam
Transducer
Flaw
Signal is scattered
For Illustration Only
This slide illustrates a flaw of the same size as the previous slide, however it is a
poor reflector and the signal is scattered. It is impossible to determine the size of
the flaw on the basis of amplitude based sizing. The scattering is exaggerated, but it
shows the concept.
-
8/6/2019 6.4 Keynote - Horsley
12/17
23
23
Tilt or Skew
Signal is scattered
For Illustration Only
This slide illustrates a flaw of the same size as the previous two slides, however it is
angled slightly off of the ideal oriented . Although it is a good reflector, the
reflected beam is not aimed directly at the receiver. It is impossible to determine
the size of the flaw on the basis of amplitude based sizing.
24
24
Inspection Challenges and Gaps AUT
Challenges
Numerous options available in design and setup of system
Accuracy in sizing adding error allowance can be a large
penalty for strain based design
How to account for human factor in assessment
Perceptions within industry
Alternative Integrity Validation (no construction hydrotest)
GAPS
Need standard procedures for design of system
Need guidance on quantifying accuracy and precision
reliability
Qualification of operators
-
8/6/2019 6.4 Keynote - Horsley
13/17
25
25
Assessment
Many new advances in weld flaw assessment:
Modeling and Experimental Capabilities
Implemented through:
Stress based standards
API 1104 Appendix A currently being revised
CSA Z662-03
Strain based standards
CSA Z662-07 in draft
DNV OS F101
26
26
Stress-Based Assessment
Challenges:
Measurement of pipe/weld/HAZ properties
Accounting for variability
Geometric variability
High/Low
Defect interaction rules poorly defined
Quantification of appropriate safety factors
Gaps:
Need better understanding of essential variables in weld
procedure
Need to quantify high/low effect
Need to quantify flaw interaction
-
8/6/2019 6.4 Keynote - Horsley
14/17
27
27
Strain-Based Design Assessment
Challenges:
All challenges from stress based design
Extra sensitivity to flaw size accuracy
Appropriate toughness measurement
Tolerable flaw size may be difficult for some welding
systems
Gaps:
All gaps from stress based design
Need low constraint small specimen toughness test
standard
Need Reliability Based Design and Assessment standard
28
A long term view.
These next few slides represent what may be possible over the next 10-15 years.
-
8/6/2019 6.4 Keynote - Horsley
15/17
29
29
Welding, Inspection and Assessment - Present
Mechanized
Welding+ AUT
Welding
Foreman
These next several slides utilize the concept of a closed-loop feedback system.
Today, on most pipeline projects utilizing mechanized welding and AUT, thewelding foreman frequently visits the AUT shack. He obtains inspection data on the
last few welds, including which side of the pipe the flaws lie, and their location with
respect to the wall thickness. He uses this information to determine which pass and
welding operator is responsible for the flaw. He investigates the problem with the
welding operator, thus completing the feedback loop.
30
30
Welding, Inspection and Assessment - Future
Mechanized
Welding+
Process
Monitoring
AUT
So-called intelligent welding systems are already in use in other industries. They
make use of data collected during the making of the weld to self-correct the process,
and thereby avoid creation of large flaws.
-
8/6/2019 6.4 Keynote - Horsley
16/17
31
31
Welding, Inspection and Assessment - Future
Mechanized
Welding+
Process
Monitoring
AUT
Flaw
Acceptance
Criteria
+
If the data from monitoring the weld process is interpreted to determine if flaws are
present, then the flaws can be compared to the flaw acceptance criteria to determine
if a repair is necessary. The AUT remains, but is only a secondary validation that
defects are not present.
32
32
Welding, Inspection and Assessment - Future
Mechanized
Welding+
Process
Monitoring
Flaw
Acceptance
Criteria
+
Process
Optimization
Prediction
of Properties
Design
Requirements
A current PRCI/DOE project is expected to deliver enhanced tools for the prediction
of microstructure and material properties on the basis of the welding process. The
inputs are material compositions, welding parameters. This could be used for
optimization of the welding process and initial material selections, to meet projectspecific pipeline design requirements. In the context of on-line assessment this
technology could be used to predict materials properties, and these could be fed into
the flaw assessment routine. Of course each of these predictions and processes are
subject to variation and scatter. However
-
8/6/2019 6.4 Keynote - Horsley
17/17
33
33
Reliability-Based Design and Assessment
Mechanized
Welding+
Process
Monitoring
Flaw
Acceptance
Criteria
+
Process
Optimization
Prediction
of Properties
Design
Requirements
Loads
Failure
Consequences
if the loads and failure consequences are defined
34
34
Reliability-Based Design and Assessment
Mechanized
Welding+
Process
Monitoring
Flaw
Acceptance
Criteria
+
Process
Optimization
Prediction
of Properties
Design
Requirements
Loads
Failure
Consequences
Probabilistic Framework
in terms of statistical variations, the entire design can be incorporated into a probabilistic framework. Weld
flaw inspection and assessment will form only one part of the overall pipeline Reliability Based Design and
Assessment.
This may seem like a bold long term objective, as it requires numerous tools to come together. Some of thetechnologies are new to the pipeline industry, and others will rely on intensive computation capabilities, i.e.
on-line prediction of properties and tolerable flaws being fed into the welding control system. However,
remember where we were fifteen years ago; it was rare to have PC on our desk, cell phones, or PDAs.
Reliability Based Design and Assessment:
Quantifies the reliability for all relevant failure conditions (limit states)
Takes into account all mitigation measures:
Pipe material and geometric e.g. grade, WT
Inspection e.g. AUT, ILI, ROW surveillance
Protection burial
Adaptable to include unique design conditions and new technology
Optimization of combined design, construction and maintenance programs to achieve acceptable
reliability/risk levels
This is a very important goal for the pipeline industry, as project economics and pipeline operation and
maintenance can be greatly improved by use of this integrated tool. It allows resources to be allocated to
where they are most effective in improving pipeline integrity and economics.